國立虎尾科技大學 |

Improving Inspector Training and Understanding Through Human-Centric Intelligence Methods for Drone-Assisted Bridge Inspection.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Improving Inspector Training and Understanding Through Human-Centric Intelligence Methods for Drone-Assisted Bridge Inspection./
作者:	Li, Yu.
面頁冊數:	1 online resource (116 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-02, Section: B.
Contained By:	Dissertations Abstracts International85-02B.
標題:	Civil engineering. -
電子資源:	click for full text (PQDT)
ISBN:	9798380141192

Improving Inspector Training and Understanding Through Human-Centric Intelligence Methods for Drone-Assisted Bridge Inspection.
Li, Yu.

Improving Inspector Training and Understanding Through Human-Centric Intelligence Methods for Drone-Assisted Bridge Inspection. - 1 online resource (116 pages)

Source: Dissertations Abstracts International, Volume: 85-02, Section: B.

Thesis (Ph.D.)--State University of New York at Stony Brook, 2023.

Includes bibliographical references

The U.S. Highway Bridge Inventory includes approximately 617,000 bridges that require inspection every two years to prevent catastrophic incidents. To improve safety and efficiency, robots such as drones have increasingly been used for bridge inspections. With a drone's assistance, inspectors can gather image and video data of a bridge including at its hard-to-reach areas. While drones can operate autonomously, human inspectors need to remain in the loop for complex tasks. The goal of using robots for bridge inspection is to enhance the capabilities of inspectors, not replace them. Therefore, inspectors need to develop the necessary skills and confidence to operate drones effectively. Moreover, transitioning from conventional to emerging inspection techniques necessitates extensive cooperation and interaction between humans and robots.My dissertation research is motivated by these needs, and it aims to innovate methods of inspector training and understanding for drone-assisted bridge inspection. To advance inspector training, I have developed a virtual reality-based system that facilitates the training and assessment of inspectors in conducting drone-assisted bridge inspections. This system provides a more immersive, interactive, and personalized training experience that can enhance inspectors' necessary skills in interacting with drones in their jobs. I also built a multi-tasking deep learning model that recognizes audio speech commands given by inspectors to guide a semi-autonomous drone in inspection. This model has a Share-Split-Collaborate architecture that allows for the sharing of the feature extractor and the splitting of subject-specific and keyword-specific features through feature projection and collaborative training. Finally, to address the lack of robustness of the audio data, I explored surface Electromyography (sEMG) sensor data as an input signal for a multi-tasking model. This approach provides an alternative input signal that is less susceptible to noise and interference, improving the reliability of the model.In conclusion, drones hold the potential to enhance the safety and efficiency of infrastructure maintenance. My research contributes to this endeavor by developing a training system and machine learning models for collaborative robot-inspector surveys of infrastructure systems.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798380141192Subjects--Topical Terms:

561339
Civil engineering.
Subjects--Index Terms:

Human robot interactionIndex Terms--Genre/Form:

554714
Electronic books.

Improving Inspector Training and Understanding Through Human-Centric Intelligence Methods for Drone-Assisted Bridge Inspection.
LDR:03825ntm a22003977 4500 001 1151828
005 20241118135834.5
006 m o d
007 cr mn ---uuuuu
008 250605s2023 xx obm 000 0 eng d
020 $a 9798380141192
035 $a (MiAaPQ)AAI30528585
035 $a AAI30528585
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Li, Yu. $3 1180731
245 1 0 $a Improving Inspector Training and Understanding Through Human-Centric Intelligence Methods for Drone-Assisted Bridge Inspection.
264 0 $c 2023
300 $a 1 online resource (116 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-02, Section: B.
500 $a Advisor: Qin, Ruwen.
502 $a Thesis (Ph.D.)--State University of New York at Stony Brook, 2023.
504 $a Includes bibliographical references
520 $a The U.S. Highway Bridge Inventory includes approximately 617,000 bridges that require inspection every two years to prevent catastrophic incidents. To improve safety and efficiency, robots such as drones have increasingly been used for bridge inspections. With a drone's assistance, inspectors can gather image and video data of a bridge including at its hard-to-reach areas. While drones can operate autonomously, human inspectors need to remain in the loop for complex tasks. The goal of using robots for bridge inspection is to enhance the capabilities of inspectors, not replace them. Therefore, inspectors need to develop the necessary skills and confidence to operate drones effectively. Moreover, transitioning from conventional to emerging inspection techniques necessitates extensive cooperation and interaction between humans and robots.My dissertation research is motivated by these needs, and it aims to innovate methods of inspector training and understanding for drone-assisted bridge inspection. To advance inspector training, I have developed a virtual reality-based system that facilitates the training and assessment of inspectors in conducting drone-assisted bridge inspections. This system provides a more immersive, interactive, and personalized training experience that can enhance inspectors' necessary skills in interacting with drones in their jobs. I also built a multi-tasking deep learning model that recognizes audio speech commands given by inspectors to guide a semi-autonomous drone in inspection. This model has a Share-Split-Collaborate architecture that allows for the sharing of the feature extractor and the splitting of subject-specific and keyword-specific features through feature projection and collaborative training. Finally, to address the lack of robustness of the audio data, I explored surface Electromyography (sEMG) sensor data as an input signal for a multi-tasking model. This approach provides an alternative input signal that is less susceptible to noise and interference, improving the reliability of the model.In conclusion, drones hold the potential to enhance the safety and efficiency of infrastructure maintenance. My research contributes to this endeavor by developing a training system and machine learning models for collaborative robot-inspector surveys of infrastructure systems.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Civil engineering. $3 561339
650 4 $a Environmental engineering. $3 557376
653 $a Human robot interaction
653 $a Human-in-the-loop
653 $a Infrastructure inspection
653 $a keyword classification
653 $a Multitask learning
653 $a Speaker recognition
655 7 $a Electronic books. $2 local $3 554714
690 $a 0543
690 $a 0775
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a State University of New York at Stony Brook. $b Civil Engineering. $3 1474494
773 0 $t Dissertations Abstracts International $g 85-02B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30528585 $z click for full text (PQDT)